Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
UDC: 656.084(560)
DOI: http://dx.doi.org/10.7708/ijtte.2014.4(3).07
RISK FACTORS AFFECTING FATAL VERSUS NON-FATAL ROAD
TRAFFIC ACCIDENTS: THE CASE OF KARS PROVINCE, TURKEY
Ali Kemal Çelik1, Ötüken Senger2
Atatürk University, Faculty of Economics and Administrative Sciences, Department of Quantitative
Methods, 25240, Erzurum, Turkey
2
Kafkas University, Faculty of Economics and Administrative Sciences, Department of Quantitative
Methods, 36100, Kars, Turkey
1
Received 6 December 2013; accepted 30 June 2014
Abstract: The aim of this paper is to determine risk factors affecting the fatal versus non-fatal
accidents in a rural region of Turkey, during 2008-2012, considering variables associated with
the individual, the environment, and the motor vehicle. A retrospective study was conducted
by obtaining the data from the traffic police road accident reports between 2008 and 2012.
A stepwise logistic regression analysis was performed to analyze the data and critical factors
that contributed significantly to fatal versus non-fatal traffic accidents. The results revealed
that the driver’s age (OR = 1.09; 90% CI = 1.05 – 1.14), clear weather (OR = 0.11; 90% CI
= 0.02 – 0.65), winter season (OR = 0.16; 90% CI = 0.03 – 0.75), straight (OR = 0.22; 90%
CI = 0.05 – 0.91) or slight road curve/bend (OR = 0.17; 90% CI = 0.04 – 0.83), the driver’s
education (OR = 0.18; 90% CI = 0.04 – 0.85) and the purpose of the vehicle (OR = 0.20;
90% CI = 0.04 – 0.94) were the significant factors affecting road traffic accidents over the
sample period.
Keywords: road traffic accidents, risk factors, logistic regression, fatal, non-fatal, Turkey.
1. Introduction
Road traffic injuries are one of the leading
causes of mortality and disability; about
1.24 million people die each year as a result
of road traffic accidents; between 20 to
50 million suffer from non-fatal injuries;
and moreover, road traffic accidents are
predicted to result in the deaths of 1.9
million people annually by 2020 (WHO,
2013a). Road traffic injuries and fatalities are
notably increasing in low-and middle income
countries; while current trends suggest that
they will become the fifth leading cause of
death by 2030, with the disparity between
high- and low-income countries. Eighty per
1
Corresponding author: [email protected]
339
cent of road traffic deaths occur in middleincome countries that comprise 72% of the
world’s population. Half of the world’s all
road traffic deaths are among motorcyclists
(23%), pedestrians (22%), and cyclists (5%),
so-called ‘vulnerable road users’, and not
surprisingly, higher proportion of them
are the citizens of low- or middle-income
countries (WHO, 2013b).
As a middle-income country in Southeastern
Europe, Turkey suffers from traffic accidents
and their negative outcomes. There has been
an overwhelmingly very large increase in the
number of road motor vehicles in Turkey
during the last ten years, such that, the
International Journal for Traffic and Transport Engineering, 2014, 4(3): 339 - 351
number of road motor vehicles has increased
by approximately 90% (Turkish National
Police, 2013a). In that period, more than 8.8
million road traffic accidents were occurred,
almost 1.1 million of them were fatal or nonfatal which affected nearly two millions
people. Specifically, 22 to 31 people were
killed, and 1,340 to 1,574 people were injured
per 100,000 vehicles (Turkish Statistical
Institute, 2012a; Turkish National Police,
2013b).
Over the past years, a great number of studies
have concentrated on manifold factors
influencing fatal and non-fatal road traffic
accidents and their injury severity. Recent
studies emphasized pre-crash human factors,
such as alcohol consumption (Kim et al.,
1995; Reynaud et al., 2002; Cummings et
al., 2006; Yannis et al., 2008; Arranz and
Gil, 2009; De Boni et al., 2013); seat-belt
or Helmet usage (Cooper and Salzberg,
1993; Kim et al., 1995; Cummings et al.,
2006); pre-crash vehicle factors such as the
type of tyres (Strandoth et al., 2012; Elvik
et al., 2013); at-crash vehicle factors such
tire blow-outs, mechanical defects of the
motor vehicles (Martin and Laumon, 2005;
Barengo et al., 2006; Alam and Spainhour,
2009); pre-crash environmental factors such
as roadway features or defects (Shankar et
al., 1995; Karlaftis and Golias, 2002; Lee and
Mannering, 2002; Chen and Chen, 2011;
Kartal et al., 2011); at crash environmental
factors such as road parameters, traffic
signs, street lights, curbs or rumble strips
(Hijar et al., 2000; Carson and Mannering,
2001; Zhou et al., 2005; Kim et al., 2007;
Bombom and Edino, 2009; Šliupas, 2009;
Pulugurtha and Bhatt, 2010; Jiang et al.,
2011; Mamčic and Sivilevičius, 2013) using
various statistical methods including logistic
regression. In some researches, age (Cooper,
1990; Abdel-Aty et al., 1998; Zhang et al.,
1998; 2000; Awadzi et al., 2008; Alam and
Spainhour, 2009), gender (Ulfarsson and
Mannering, 2004; Islam and Mannering,
2006; Bener and Crundall, 2008; Majdzadeh
et al., 2008), or both factors (Holubowycz
et al., 1994; Massie et al., 1995; Chipman,
1995; Glendon et al., 1996; Laapotti and
Keskinen, 1998; Jones and Jørgensen, 2003;
Kaplan and Prato, 2012) are examined as
specific risk factors.
2. Materials and Methods
2.1. General Traffic Information about
Kars City
Kars is a city located in Northeast Turkey
with a population of 304,821; an area of
10,127 square kilometers; and an altitude
of 1768 meters including its province, 8
districts, and 382 villages (Governor of Kars,
2013). Fig. 1 depicts the road traffic map
of the city. Kars city covers 736 kilometers
of network length, 675 kilometers of total
road including 499 kilometers of national,
and 176 kilometers of provincial highway
(Republic of Turkey General Directorate
of Highways, 2013). As of December, 31,
2012, there are 94,797 registered drivers
and 35,216 motor vehicles in Kars including
14,927 tractors, 9,257 automobiles, 5,854
single-unit trucks, 1,870 trucks, 1,641
minibuses, 1,126 motorcycles, 417 buses,
and 129 special-purposed vehicles (Turkish
Statistical Institute, 2012b; Turkish National
Police, 2013c).
340
Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
Fig. 1.
Traffic Highway Map of Kars
Source: General Directorate of Highways
The traffic accident data used in this study
were obtained from road traffic accident
reports of Traffic Services Branch Office
and Regional Traffic Control Branch Office
under the responsibility of Kars Provincial
Police Department. The corresponding
data involved 765 fatal and non-fatal road
traffic accidents which occurred on Kars
city and its central districts’ roads during
2008-2012. The paper used a simple
random sampling method to investigate
the data of 765 traffic police-reported
accidents, while the data were transformed
and coded to a convenient computer-ready
form. A fatal-injury accident is defined as
an accident in which at least one person
(driver, passenger or pedestrian) was killed
at crash. A non-fatal injury accident refers
to an accident in which at least one person
was suffered injury but no fatalities were
occurred.
2.2. Methods
Most studies have several explanator y
variables, which they may be continuous as
well as categorical. In that context, a good-
341
fitting model will evaluate their effects on
response variables, will include relevant
interactions, and will provide smoothed
estimates of response probabilities (Agresti,
2002). While an observation taking one of
two possible forms on each individual is
supposed, and if for the ith individual, this
observation is represented by a random
variable, y; then without loss of generality
code, two possible values of y by 1 and 0 can
be defined as follows (Eq. (1)):
E (y) = Prob (y = 1) = π, Prob (y = 0) = 1 – π
(1)
In Eq. (1), y = 1 and y = 0 are usually called
a ‘success’, and ‘failure’, respectively, and
such observations are considered as ‘binary’
(Cox, 1970).
For a binary response Y and a quantitative
independent variable X, when π(x) denotes
the success probability, the logistic regression
model has linear form for the logit of the
corresponding probability as follows (Eq.
(2)):
International Journal for Traffic and Transport Engineering, 2014, 4(3): 339 - 351
g(x) = log
(2)
The left-hand side of Eq. (2) is called the logodds ratio (OR), and the odds of response
variable can be derived (Agresti, 1996):
= exp(
)
(3)
The odds initially indicate how often
something happens relative to how often it
does not happen (Long, 1997). The log of the
odds transformation performs the conversion
of the probability estimates to a continuous
unbounded variable. This variable will
become the dependent variable in a linear
model with the categorical definitions as
independent variables (Hanushek and
Jackson, 1977). In Eq. (3), for every oneunit increase in x, the odds will increase
multiplicatively by e β , namely, the odds
at level x + 1 will equal to the odds at x
multiplied by eβ Specifically, when β = 0, eβ
will be equal to 1, and the odds do not change
as x changes (Agresti, 1996).
Stepwise logistic regression procedure
enables a useful and effective data analysis
tool and employing a stepwise procedure
can provide a fast and effective means to
screen a large number of variables, and to
fit a number of logistic regression equations
simultaneously. Any stepwise procedure
for selection or deletion of variables from
a model is based on a statistical algorithm
that checks for the importance of variables
(Hosmer and Lemeshow, 2000). In this
study, because the dependent variable is
a binar y or dichotomous variable, the
stepwise logistic regression is an appropriate
technique, which is developed to predict a
binary dependent variable as a function of
independent variables. This technique is
frequently used in road safety where the
dependent variable is binary (Chipman,
1995; Zhang et al., 1998, 2000; Shon and
Shin, 2001; Al-Ghamdi, 2002; Reynaud et
al., 2002; Jones and Jørgensen, 2003; Yau,
2004; Sze and Wong, 2007; Awadzi et al.,
2008; Majdzadeh et al., 2008; Tay et al.,
2008; Kartal et al., 2011; De Boni et al., 2013;
Drucker et al., 2013).
3. Results
This study investigates the effects of twentyone factors on fatal versus non-fatal road
traffic accidents in Kars during 2008-2012,
as shown in Table 1, which also indicates
the mean, standard deviation. In this study,
except for age, all variables were qualitative,
where age was interpreted as a continuous
variable. In order to clarify the results, most
of the variables were defined as dummy
variables, which take only 0 or 1. Gender,
nationality and seat belt factors were also
omitted because of the dominancy of male
and Turkish drivers, respectively. The
traffic police mentioned that they had any
opportunity to recognize the seat-belt status
of the driver at crash, and they have coded
the status as ‘unknown’, so security factors
were omitted from the model. Similarly,
because pedestrians’ information was unclear
in the reports, it was excluded and only
drivers’ information was included.
342
Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
Table 1
Descriptive Statistics of Variables
Variable
Description
Mean
S.D.
Accident type (Q)
Dependent variable (non-fatal = 0; fatal = 1)
0.029
0.167
(1) Season (Q)
Spring = 1; Otherwise = 0
Summer = 1; Otherwise = 0
Autumn = 1; Otherwise = 0
Winter = 1; Otherwise = 0
0.191
0.277
0.320
0.212
0.393
0.448
0.467
0.409
(2) Hour (Q)
00:00-05:59 = 1; Otherwise = 0
06:00-11:59 = 1; Otherwise = 0
12:00-17:59 = 1; Otherwise = 0
18:00-23:59 = 1; Otherwise = 0
0.065
0.261
0.379
0.294
0.247
0.440
0.485
0.456
(3) Weather (Q)
Clear = 1; Otherwise = 0
Cloudy/Rainy/Foggy = 1; Otherwise = 0
Snowy/Stormy = 1; Otherwise = 0
0.668
0.264
0.068
0.471
0.441
0.252
(4) Time of day (Q)
Night time = 1; Day time = 0
0.357
0.479
(5) Site (Q)
Formal settlement = 1; Informal settlement = 0
0.601
0.506
(6) Occurance of accident (Q)
Overturn/Falling person, animal or object from the vehicle/
Skidding from the road = 1; Collision = 0
0.461
0.499
(7) Number of vehicle(s) (Q)
Single vehicle = 1; Multi-vehicle = 0
0.361
0.481
(8) Divided highway (Q)
Divided highway = 1; Undivided highway = 0
0.659
0.474
(9) Road surface (Q)
Dry/Dusty =1; Otherwise = 0
Wet/Puddle/Oil on the pavement = 1; Otherwise = 0
Snowy/Icy = 1; Otherwise = 0
0.677
0.152
0.171
0.468
0.359
0.377
(10) Direction of the road (Q)
Two-way = 1; One-way = 0
0.797
0.402
(11) Working on the road (Q)
Available = 1; not available = 0
0.103
0.305
(12) Road lane line (Q)
Available = 1; not available = 0
0.604
0.513
(13) Horizontal route (Q)
0.782
0.180
0.038
0.252
0.385
0.191
(14) Vertical route (Q)
Straight road = 1; otherwise = 0
Slight road curve/Bend = 1; Otherwise = 0
Hard road curve/Bend with or without rail = 1;
Otherwise = 0
Slight slope/Steep slope/Over the hill slope = 1; No slope = 0;
0.363
0.481
(15) Side walk (Q)
Available = 1; not available = 0
0.207
0.405
(16) Age (C)
Age of the driver
36.24
10.88
(17) Education level (Q)
Education level of the driver
Primary education = 1; otherwise = 0
Secondary education = 1; otherwise = 0
Higher education = 1; otherwise = 0
0.492
0.310
0.199
0.500
0.463
0.399
(18) Alcohol use (Q)
Drunk = 1; otherwise = 0
(19) Type of vehicle (Q)
Auto = 1; otherwise = 0
Single-unit truck/truck/bus = 1; otherwise = 0
Others (i.e. motorcycle, tractor, ambulance, military vehicle)
0.142
0.350
(20) Purpose of vehicle (Q)
Private = 1; otherwise = 0
Commercial = 1; otherwise = 0
Others (i.e. security, military, agricultural) = 1;
otherwise = 0
0.516
0.384
0.099
0.500
0.487
0.299
0.724
0.184
0.081
0.447
0.388
0.273
(21) Contributing circumstance (Q)
Speeding = 1; otherwise = 0
Inattention and negligence = 1; otherwise = 0
Others (i.e. violating transition rule, rear-end collision/
improper lane changing/lane rape, lack of tecnical
requirements) = 1; otherwise = 0
0.545
0.184
0.325
0.498
0.388
0.469
(Q) = Qualitative variable
(C) = Continuous variable
(Q ) = Qualitative variable
(C) = Continuous variable
343
International Journal for Traffic and Transport Engineering, 2014, 4(3): 339 - 351
Table 2
Logistic Regression Estimation Results of Risk Factors Affecting Fatal vs Non-Fatal Road Traffic
Accidents
Variable
OR
Std. Err.
Z
p-value
(1) Season (base Spring)
Summer
Autumn
Winter
(2) Hour (base 00:00 – 05:59)
06:00 – 11:59
12:00 – 17:59
18:00 – 23:59
(3) Weather (base Snowy/Stormy)
Clear
Cloudy/Rainy/Foggy
(4) Time of day
Night Time
(5) Site
Formal Settlement
(6) Occurance of accident
Overturn/Falling person, animal or object from the vehicle/
Skidding from the road
(7) Number of vehicle(s)
Single vehicle
(8) Divided highway
Divided Highway
(9) Road surface (base Wet/Oil on the Pavement)
Dry/Dusty
Snowy/Icy
(10) Direction of the road
Two-way
(11) Work on the road
Available
(12) Road lane line
Available
(13) Horizontal route (base Hard road curve/Bend with or
without rail)
Straight road
Slight road curve/bend
(14) Vertical route
Slight slope/Steep slope/Over the hill slope
(15) Side walk
Available
(16) Age
(17) Education level(base Higher Education)
Primary education
Secondary education
(18) Alcohol use
Drunk
(19) Type of vehicle (base Others)
Auto
Single-unit truck/truck/bus
(20) Purpose of vehicle (base Others)
Private
Commercial
(21) Contributing circumstance (base Others)
Speeding
Inattention and negligence
Number of observations = 765
Log likelihood = -71.163
Logistic regression X2 = 57.18
Prob > X2 = 0.0056
Pseudo R2 = 0.2866
0.41
0.38
0.16
0.307
0.263
0.148
-1.19
-1.40
-1.95
0.234
0.162
0.052***
0.93
0.32
0.43
1.325
0.427
0.442
-0.05
-0.86
-0.82
0.959
0.392
0.411
0.11
0.50
0.120
0.503
-2.04
-0.69
0.041**
0.490
2.21
2.137
0.82
0.413
0.99
0.752
-0.01
0.995
0.34
0.276
-1.33
0.184
1.67
1.717
0.50
0.620
2.13
1.453
1.11
0.266
1.55
0.40
1.611
0.509
0.43
-0.72
0.670
0.472
0.41
0.319
-1.15
0.252
0.66
0.760
-0.36
0.720
1.87
1.103
1.07
0.286
0.22
0.17
0.189
0.165
-1.76
-1.85
0.078***
0.065***
2.71
1.731
1.56
0.118
0.19
1.09
00
0.287
01
0.027
-1.10
3.67
0.273
0.000*
-1.25
-1.82
0.210
0.069***
0.71
0.476
0.80
1.26
0.422
0.207
-1.71
-1.23
0.088***
0.217
0.75
0.82
0.454
0.411
0.39
0.18
1.80
2.34
3.66
0.20
0.27
2.15
2.39
0.293
0.171
1.473
2.471
3.767
0.188
0.286
2.208
2.533
[90% C.I.]
0.12 – 1.40
0.12 – 1.19
0.03 – 0.75
0.09 – 9.72
0.04 – 2.85
0.08 – 2.33
0.02 – 0.65
0.09 – 2.63
0.45 – 10.9
0.29 – 3.45
0.09 – 1.29
0.31 – 9.07
0.70 – 6.54
0.28 – 8.55
0.05 – 3.24
0.11 – 1.48
0.10 – 4.37
0.71 – 4.93
0.05 – 0.91
0.04 – 0.83
0.95 – 7.75
0.02 – 2.30
1.05 – 1.14
0.12 – 1.34
0.04 – 0.85
0.47 – 6.92
0.41 – 13.3
0.67 – 19.8
0.04 – 0.94
0.05 – 1.55
0.40 – 11.6
0.42 – 13.7
1
*Significant at 1% level **Significant at 5% level ***Significant at 5% level
344
Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
Table 2 indicates the estimation results for
the final stepwise logistic regression model.
Based on p-values, 7 variables from 24 factors
were found to be significant or marginally
significant. Number of observations was 765,
with a log likelihood value of -71.163 and
chi-square was equal to 57.18. As shown in
Table 2, fatal road traffic accidents in Kars
were more likely to occur with respect to
the driver’s age (OR = 1.09; 90% CI = 1.05
– 1.14), where the corresponding variable
was marginally significant. Fatal road
traffic accidents were 89% less likely to
occur when the weather was clear (OR =
0.11; 90% CI = 0.02 – 0.65). Additionally,
fatal traffic accidents were 84% less likely
to occur in winter (OR = 0.16; 90% CI =
0.03 – 0.75). Kars has a continental climate
and the weather is usually rainy or snowy
throughout the year, even in the spring and
summer, for that reason vehicles are also
well-prepared for the negative effects of
the weather conditions. In that context, the
impacts of season and weather factors on the
occurrence of fatal and non-fatal road traffic
accidents are not surprising.
The results showed that horizontal route
factor had an impact on the probability of
fatal and non-fatal road traffic accidents.
Herein, when an accident was occurred at
345
straight road (OR = 0.22; 90% CI = 0.05 –
0.91) or slight road curve/bend (OR = 0.17;
90% CI = 0.04 – 0.83), it was less likely
to occur a fatal accident. Furthermore,
the driver’s education level had impact
on the probability of occurrence of fatal
road traffic accidents. They were less
likely to occur when the driver’s education
level was secondary (OR = 0.18; 90% CI
= 0.04 – 0.85). The results also showed
the significant impact of the purpose of
the vehicle (OR = 0.20; 90% CI = 0.04 –
0.94) on fatal versus non-fatal road traffic
accidents. One unit increase in private
vehicle variable had a 0.20 decreasing
impact on the likelihood of fatal accidents
against non-fatal accidents.
T he va r ia nce i n f lat ion fac tor (V I F)
faci l itates to measure how much
multicollinearity has increased the variance
of a slope estimate (Stine, 1995). In practice,
since VIF is less than 10, the researchers
can suggest that no variables cause the
multicollinearity problem in the analysis.
In Table 3, V IF values of independent
variables used in this study were presented to
ensure that there was not a multicollinearity
problem among these independent variables,
where all VIF values, including the mean
VIF, are less than 10.
International Journal for Traffic and Transport Engineering, 2014, 4(3): 339 - 351
Table 3
Variance Inflation Factors of Independent Variables
Variable
(1) Season
Summer
Autumn
Winter
(2) Time
06:00-11:59
12:00-17:59
18:00-23:59
(3) Weather
Clear
Cloudy/Rainy/Foggy
(4) Time of day
Night Time
(5) Site
Formal Settlement
(6) Occurance of accident
Overturn/Falling person, animal or object from the vehicle/Skidding from the road
(7) Number of vehicle(s)
Single vehicle
(8) Divided highway
Divided Highway
(9) Road surface (base Wet/Oil on the Pavement)
Dry/Dusty
Snowy/Icy
(10) Direction of the road
Two-way
(11) Work on the road
Available
(12) Road lane line
Available
(13) Horizontal route (base Hard road curve/Bend with or without rail)
Straight road
Slight road curve/bend
(14) Vertical route
Slight slope/Steep slope/Over the hill slope
(15) Side walk
Available
(16) Age
(17) Education level(base Higher Education)
Primary education
Secondary education
(18) Alcohol use
Drunk
(19) Type of vehicle (base Others)
Auto
Single-unit truck/truck/bus
(20) Purpose of vehicle (base Others)
Private
Commercial
(21) Contributing circumstance (base Others)
Speeding
Inattention and negligence
Mean VIF
VIF
1/VIF
1.94
1.97
2.26
0.515
0.509
0.442
6.87
7.24
4.35
0.146
0.138
0.230
5.70
5.14
0.175
0.195
3.10
0.323
2.25
0.445
2.41
0.415
3.57
0.280
1.75
0.571
3.70
2.87
0.270
0.348
1.61
0.622
1.16
0.863
1.53
0.655
5.34
4.89
0.187
0.204
1.51
0.664
2.05
1.15
0.489
0.871
2.17
1.95
0.460
0.513
1.24
0.806
4.47
4.18
0.224
0.239
3.59
3.33
0.279
0.300
3.68
2.04
0.272
0.489
3.12
346
Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
Table 4
Goodness of Fit Statistics
Number of observations
Number of covariate patterns
Pearson X2(724)
Hosmer-Lemeshow X2(8)
Prob > X2
765
758
405.88
2.16
0.98
4. Conclusion
The Eastern regions of Turkey are relatively
underdeveloped and exposed to chronic
economic and financial problems. In recent
years, automobiles and various types of
trucks had also produced in the occurrence of
fatal and/or non-fatal road traffic accidents.
The great number of motor vehicles involved
in the road traffic is one of the common
issues of traffic safety in Turkey and the
rural regions. Otherwise, Kars city suffers
from unfavorable weather conditions; where
average number of rainy days exceeds ten
days in a month with average total amount
of rainfall ranging between 20.3 and 77.2
k ilograms/square meters in a year. In
addition, as a result of continental climate,
Kars city is one of the coldest cities in Turkey,
such that the number of snowy days exceeds
120 days (Turkish State Meteorological
Service, 2013). In that context, weather
conditions tend to negatively affect fatal
and/or non-fatal road traffic accidents.
By t he cou r te s y of v a r iou s reg ion a l
development projects and Turkish governors’
dominancy, roadways and the road traffic
have been remarkably improved during the
last five years in Northeast Turkey. However,
there are still significant regional disparities
between the West and the East. The Eastern
regions survive with economic difficulties
that inherently have an impact on the road
traffic. Despite optimistic development
347
efforts, the corresponding territories still
need qualified roadways to overcome the lack
of countrywide essential transition. Decision
makers should concentrate on more durable
roadway construction ventures, namely,
asphalts and other traffic infrastructures do
not negatively affected by weather conditions
in a little while. From a different perspective,
more educated t ra f f ic person nel a nd
conscious drivers will be probably helpful to
create a more favorable traffic environment.
According to the results of this study, age
variable was highlighted as a marginally
significant risk factor affecting fatal road
traffic accident against non-fatal. Since driver’s
age particularly increases the probability of
occurrence for fatal road traffic accidents,
the crucial roles of both policy makers and
young drivers’ parents revisit. In this sense,
parents may preemptively avoid adolescents
to drive before permitted by the relevant
legislation. Similarly, traffic security laws
may be re-arranged by deterrent regulations
to preclude the satisfaction of inexperienced
young drivers. Since most of the traffic users
properly respect traffic safety regulations,
a n i mpor t a nt step may be su r pa ssed .
Nevertheless, decision makers should draw
attention to the poor reporting of road traffic
accidents. For instance, risk factors such as seat
belt and alcohol use of the traffic users should
be more efficiently reported to determine the
virtual reasons of the road traffic accidents
more precisely. On the other hand, road traffic
accidents in Turkey are considered as fatal
accidents when one of the traffic users is fatally
injured when the accident occurs and the
elapsed time period in the hospital is ignored
in contrast to the procedure applied in many
developed countries. In that circumstance,
the number of actual fatally injured traffic
users is not properly illustrated. Emerging
countries should adapt their traffic legislation
International Journal for Traffic and Transport Engineering, 2014, 4(3): 339 - 351
to developed countries to reach better traffic
reporting standards and thus achieving an
improved traffic safety policy.
Acknowledgements
The authors would like to thank Kars Provincial
Traffic Police Department, especially Traffic
Services Branch and Regional Traffic Control
Branch Offices’ chief superintendents and all
the staff for their supports, special interests,
and courtesy in charge of collecting nonpersona l ized data a nd adv ice on data
management that made this study possible.
Awadzi, K.D.; Classen, S.; Hall, A.; Duncan, R.P.;
Garvan, C.W. 2008. Predictors of injury among younger
and older adults in fatal motor vehicle crashes, Accident
Analysis & Prevention. DOI: http://dx.doi.org/10.1016/j.
aap.2008.07.010, 40(6): 1804-1810.
Barengo, N.C.; Mkamba, S.M.; Mshana, M.; Miettola,
J. 2006. Road traffic accidents in Dar-es-Salaam,
Tanzania during 1999 and 2001, International Journal of
Injury Control and Safety Promotion. DOI: http://dx.doi.
org/10.1080/15660970500036713, 13(1): 52-54.
References
Bener, A.; Crundall, D. 2008. Role of gender and
driver behaviour in road traffic crashes, International
Journal of Crashwor thiness. DOI: ht t p://d x .doi.
org/10.1080/13588260801942684, 13(3): 331-336.
Abdel-Aty, M.A.; Chen, C.L.; Schott, J.R. 1998. An
assessment of the effect of driver age on traffic accident
involvement using log-linear models, Accident Analysis
& Prevention. DOI: http://dx.doi.org/10.1016/S00014575(98)00038-4, 30(6): 851-861.
Bombom, L.S.; Edino, M.O. 2009. Road profiling of
traffic accidents in Jos, Nigeria, 1995-1999, International
Journal of Injury Control and Prevention. DOI: http://
dx.doi.org/10.1080/17457300903024186, 16(3):
169-174.
Agresti, A. 1996. An Introduction to Categorical Data
Analysis. John Wiley & Sons, Inc. USA. 400 p.
Carson, J.; Mannering, F. 2001. The effect of ice warning
signs on ice-accident frequencies and severities, Accident
Analysis & Prevention. DOI: http://dx.doi.org/10.1016/
S0001-4575(00)00020-8, 33(1): 99-109.
Agresti, A. 2002. Categorical Data Analysis. John Wiley
& Sons, Inc. USA. 744 p.
Alam, B.M.; Spainhour, L.K. 2009. Contributing factors
for young at fault drivers in fatal traffic crashes in Florida,
Journal of Transportation Safety & Security. DOI: http://
dx.doi.org/10.1080/19439960902735444, 1(2): 152-168.
Al-Ghamdi, A.S. 2002. Using logistic regression to
estimate the influence of accident factors on accident
severity, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/S0001-4575(01)00073-2, 34(6):
729-741.
Arranz, J.M.; Gil, A.I. 2009. Traffic accidents, deaths
and alcohol consumption, Applied Economics. DOI: http://
dx.doi.org/10.1080/00036840701222652, 41(20):
2583-2595.
Chen, F.; Chen, S. 2011. Injury severities of truck
drivers in single- and multi-vehicle accidents on rural
highways, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/j.aap.2011.03.026, 43(5): 16771688.
Chipman, M.L. 1995. Risk factors for injury: Similarities
and differences for traffic crashes and other causes,
Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/0001-4575(95)00013-P, 27(5): 699-706.
Cooper, P.J. 1990. Differences in accident characteristics
among elderly drivers and between elderly and middleaged drivers, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/0001-4575(90)90044-L, 22(5):
499-508.
348
Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
Cooper, P.J.; Salzberg, P. 1993. Safety restraint
usage in fatal motor vehicle crashes, Accident Analysis
& Prevention. DOI: http://dx.doi.org/10.1016/00014575(93)90097-G, 25(1): 67-75.
Cox, D.R. 1970. Analysis of Binary Data. Chapman &
Hall Ltd. UK. 142 p.
Cummings, P.; Rivara, F.P.; Olson, C.M.; Smith, K.M.
2006. Changes in traffic crash mortality rates attributed
to use of alcohol, or lack of a seat-belt, air bag, motorcycle
helmet, or bicycle helmet, United States, 1982-2001,
Injury Prevention. DOI: http://dx.doi.org/10.1136/
ip.2005.010975, 12(3): 148-154.
De Boni, R.; Cruz, O.G.; Weber, E.; Hasenack, H.;
Lucatelli, L.; Duarte, P.; Gracie, R.; Pechansky, F.;
Bastos, F.I. 2013. Traffic crashes and alcohol outlets in
a Brazilian state capital, Traffic Injury Prevention. DOI:
http://dx.doi.org/10.1080/15389588.2012.690164
, 14(1): 86-91.
Drucker, C.; Gerberich, S.G.; Manser, M.P.; Alexander,
B.H.; Church, T.R.; Dyan, A.D.; Becic, E. 2013. Factors
associated with civilian drivers involved in crashes
with emergency vehicles, Accident Analysis & Prevention.
DOI: http://dx.doi.org/10.1016/j.aap.2013.02.035,
55: 116-123.
Elvik, R.; Fridstrøm, L.; Kaminska, J.; Meyer, S.F.
2013. Effects on accidents of changes in the use of
studded tyres in major cities in Norway: A long-term
investigation, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/j.aap.2013.02.004, 54: 15-25.
Glendon, A.I.; Dorn, L.; Davies, D.R.; Matthews, G.;
Taylor, R.G. 1996. Age and gender differences in perceived
accident likelihood and driver competences, Risk Analysis.
DOI: http://dx.doi.org/10.1111/j.1539-6924.1996.
tb00826.x, 16(6): 755-762.
Governor of Kars. 2013. Geographical framework of
Kars. Available from Internet: <http://www.kars.gov.
tr/kars_cografi_yapi.html>.
349
Hanushek, E.A.; Jackson, J.E. 1977. Statistical Methods
for Social Scientists. Academic Press, Inc. USA. 374 p.
Hijar, M.; Carrillo, C.; Flores, M.; Anaya, R.; Lopez, V.
2000. Risk factors in highway accidents: A case control
study, Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/S0001-4575(99)00116-5, 32(5): 703-709.
Holubowycz, O.T.; Kloeden, C.N.; McLean, A.J. 1994.
Age, sex, and blood alcohol concentration of killed and
injured drivers, riders, and passengers, Accident Analysis
& Prevention. DOI: http://dx.doi.org/10.1016/00014575(94)90039-6, 26(4): 483-492.
Hosmer, D.W.; Lemeshow, S. 2000. Applied Logistic
Regression. John Wiley & Sons, Inc. USA. 375 p.
Islam, S.; Mannering, F. 2006. Driver aging and its effect
on male and female single-vehicle accident injuries:
Some additional evidence, Journal of Safety Research.
DOI: http://dx.doi.org/10.1016/j.jsr.2006.04.003,
37(3): 267-276.
Jiang, X.; Yan, X.; Huang, B.; Richards, S.H. 2011.
Influence of curbs on traffic crash frequency on highspeed roadways, Traffic Injury Prevention. DOI: http://
dx.doi.org/10.1080/15389588.2011.578285, 12(4):
412-421.
Jones, A.P.; Jørgensen, S.H. 2003. The use of multilevel
models for the prediction of road accident outcomes,
Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/S0001-4575(01)00086-0, 35(1): 59-69.
Kaplan, S.; Prato, C.G. 2012. Associating crash
avoidance maneuvers with driver attributes and accident
characteristics: A mixed logit model approach, Traffic
Injury Prevention. DOI: http://dx.doi.org/10.1080/153
89588.2011.654015, 13(3): 315-326.
Karlaftis, M.G.; Golias, I. 2002. Effects of road geometry
and traffic volumes on rural roadway accident rates,
Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/S0001-4575(01)00033-1, 34(3): 357-365.
International Journal for Traffic and Transport Engineering, 2014, 4(3): 339 - 351
Kartal, M.; Kutlar, A.; Begen, A. 2011. Logistic regression
analysis of risk factors affecting traffic accidents: Study
of Sivas, Kayseri and Yozgat (trans. from Turkish), The
International Journal of Economic and Social Research, 7(2): 45-68.
Kim, K.; Nitz, L.; Richardson, J.; Li, L. 1995. Personal
and behavioral predictors of automobile crash and injury
severity, Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/0001-4575(95)00001-G, 27(4): 469-481.
Kim, J.K.; Kim, S.; Ulfarsson, G.F.; Porello, L.A. 2007.
Bicyclist injury severities in bicycle-motor vehicle
accidents, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/j.aap.2006.07.002, 39(2): 238-251.
Laapotti, S.; Keskinen, E. 1998. Differences in fatal
loss-of-control accidents between young male and female
drivers, Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/S0001-4575(97)00121-8, 30(4): 435-442.
Lee, J.; Mannering, F. 2002. Impact of roadside features
on the frequency and severity of run-off-roadway
accidents: an empirical analysis, Accident Analysis &
Prevention. DOI: http://dx.doi.org/10.1016/S00014575(01)00009-4, 34(2): 149-161.
Long, S. 1997. Regression Models for Categorical and Limited
Dependent Variables. SAGE Publications. UK. 297 p.
Majdzadeh, R.; Khalagi, K.; Naraghi, K.; Motevalian,
A.; Eshraghian, M.R. 2008. Determinants of traffic
injuries in drivers and motorcyclists involved in an
accident, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/j.aap.2007.03.019, 40(1): 17-23.
Mamčic, S.; Sivilevičius, H. 2013. The analysis of traffic
accidents on Lithuanian regional gravel roads, Transport.
DOI: http://dx.doi.org/10.3846/16484142.2013.7828
94, 28(1): 108-115.
Martin, J.; Laumon, B. 2005. Tire blow-outs and
motorway accidents, Traffic Injury Prevention. DOI: http://
dx.doi.org/10.1080/15389580490510174, 6(1): 53-55.
Massie, D.L.; Campbell, K.L.; Williams, A.F. 1995.
Traffic accident involvement rates by driver age and
gender, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/0001-4575(94)00050-V, 27(1):
73-87.
Pulugurtha, S.S.; Bhatt, J. 2010. Evaluating the role of
weaving section characteristics and traffic on crashes
in weaving areas, Traffic Injury Prevention. DOI: http://
dx.doi.org/10.1080/15389580903370039, 11(1): 104113.
Republic of Turkey General Directorate of Highways.
2013. Road Network of 18 th Regional Directorate.
Available from Internet: <http://www.kgm.gov.tr>.
Reynaud, M.; Le Breton, P.; Gilot, B.; Vervialle,
F.; Falissard, B. 2002. Alcohol is the main factor in
excess traffic accident fatalities in France, Alcoholism:
Clinical and Experimental Research. DOI: http://dx.doi.
org/10.1111/j.1530-0277.2002.tb02490.x, 26(12):
1833-1839.
Shankar, V.; Mannering, F.; Barfield, W. 1995. Effect
of roadway geometrics and environmental factors on
rural freeway accident frequencies, Accident Analysis
& Prevention. DOI: http://dx.doi.org/10.1016/00014575(94)00078-Z, 27(3): 371-389.
Shon, S.Y.; Shin, H. 2001. Pattern recognition for road
traffic accident severity in Korea, Ergonomics. DOI:
http://dx.doi.org/10.1080/00140130120928, 44(1):
107-117.
Šliupas, T. 2009. The impact of road parameters and the
surrounding area on traffic accidents, Transport. DOI:
http://dx.doi.org/10.3846/1648-4142.2009.24.42-47,
24(1): 42-47.
Stine, R.A. 1995. Graphical interpretation of variance
inflation factors, The American Statistician. DOI: http://
dx.doi.org/10.1080/00031305.1995.10476113, 49(1):
53-56.
350
Çelik A. K. et al. Risk Factors Affecting Fatal Versus Non-Fatal Road Traffic Accidents: The Case of Kars Province, Turkey
Strandroth, J.; Rizzi, M.; Olai, M.; Lie, A.; Tingvall, C.
2012. The effects of studded tires on fatal crashes with
passenger cars and the benefits of electronic stability
control (ESC) in Swedish winter driving, Accident
Analysis & Prevention. DOI: http://dx.doi.org/10.1016/j.
aap.2011.11.005, 45: 50-60.
Sze, N.N.; Wong, S.C. 2007. Diagnostic analysis of the
logistic model for pedestrian injury severity in traffic
crashes, Accident Analysis & Prevention. DOI: http://dx.doi.
org/10.1016/j.aap.2007.03.017, 39(6): 1267-1278.
Tay, R.; Rifaat, R.M.; Chin, H.C. 2008. A logistic model
of the effects of roadway, environmental, vehicle, crash
and driver characteristics on hit-and-run crashes, Accident
Analysis & Prevention. DOI: http://dx.doi.org/10.1016/j.
aap.2008.02.003, 40(4): 1330-1336.
Turkish State Meteorological Service. 2013. Official
Statistics of Provinces and Districts. Available from
Internet: <http://www.mgm.gov.tr/veridegerlendirme/
il-ve-ilceler-istatistik.aspx?m=KARS>.
Turkish Statistical Institute. 2012a. Traffic accident
statistics (Road) 2011. Available from Internet: <www.
tuik.gov.tr/IcerikGetir.do?istab_id=70>.
Turkish Statistical Institute. 2012b. Road motor vehicle
statistics, News Bulletin, October 2012. Available from
Internet: <http://www.tuik.gov.tr/PreHaberBultenleri.
do?id=10965>.
Turkish National Police. 2013a. Burden of the type of
motor vehicles. Available from Internet: <http://www.
trafik.gov.tr/Sayfalar/Istatistikler/aracsurucu4.aspx>.
Turkish National Police. 2013b. General Traffic Accident
Statistics. Available from Internet: <http://www.trafik.
gov.tr/Sayfalar/Istatistikler/Genel-Kaza.aspx>.
Turkish National Police. 2013c. Burden of the number
of drivers with respect to provinces. Available from
I nter net: <htt p://w w w.tra f i k .gov.tr/Say fa lar/
Istatistikler/aracSurucu1.aspx>.
351
Ulfarsson, G.F.; Mannering, F.L. 2004. Differences in
male and female injury severities in sport-utility vehicle,
minivan, pickup and passenger car accidents, Accident
Analysis & Prevention. DOI: http://dx.doi.org/10.1016/
S0001-4575(02)00135-5, 36(2): 135-147.
World Health Organization. 2013a. Road traffic injuries.
Fact Sheet N°358. Available from Internet: <http://
www.who.int/mediacentre/factsheets/fs358/en/>.
World Health Organization. 2013b. Global status report
on road safety 2013: Supporting a decade of action, Geneva,
Switzerland: World Health Organization.
Yannis, G.; Papadimitrou, E.; Antoniou, C. 2008. Impact
of enforcement on traffic accidents and fatalities: A
multivariate multilevel analysis, Safety Science. DOI:
http://dx.doi.org/10.1016/j.ssci.2007.01.014, 46(5):
738-750.
Yau, K.K.W. 2004. Risk factors affecting the severity of
single vehicle traffic accidents in Hong Kong, Accident
Analysis & Prevention. DOI: http://dx.doi.org/10.1016/
S0001-4575(03)00012-5, 36(3): 333-340.
Zhang, J.; Fraser, S.; Lindsay, J.; Clarke, K.; Mao, Y.
1998. Age-specific patterns of factors related to fatal
motor vehicle traffic crashes: Focus on young and elderly
drivers, Public Health. DOI: http://dx.doi.org/10.1038/
sj.ph.1900485, 112(5): 289-295.
Zhang, J.; Lindsay, J.; Clarke, K.; Robbins, G.; Mao,
Y. 2000. Factors affecting the severity of motor
vehicle traffic crashes involving elderly drivers in
Ontario, Accident Analysis & Prevention. DOI: http://
dx.doi.org/10.1016/S0001-4575(99)00039-1, 32(1):
117-125.
Zhou, J.; Zhao, X.; Jiang, Z.; Zhu, S.; Zhou, J. 2005.
Investigation on road traffic safety of the mountain
areas in Southwest China based on Wulong county,
Chongqing Municipality, Traffic Injury Prevention. DOI:
http://dx.doi.org/10.1080/15389580590931707, 6(2):
193-196.